[EM] The new method and the example

[MIKE OSSIPOFF]

Forrest--

When you posted that example, you told how IRV chooses, and how a
new method chooses.

I admit that haven't studied what it would be necessary to study to
understand the new method's definition. And of course without understanding 
how it works, it would be impossible to say what its
properties would be.

But sometimes something can be inferred from what goes into the
black box, and what comes out.

In your example, for instance, in what way is A a better choice than
B? B is overall more highly-rated in the rankings, and so B is the
Borda winner. A & B share the same magnitude of defeats, and the
only difference is that one of B's victories is stronger than those of
A.

I'm not saying that Borda's winner is always the best. If voting hasn't been 
sincere, or if we have reason to believe that rating differences
aren't uniform in rankings, or if there are factions running different
numbers of clones or other very similar candidates, for instance.
Or when Borda's choice avoidably violates majority rule.


But those conditions don't exists in your example, and the fact that
B is the Borda winner, the overall highest-ranked candidate, especially 
underlines the question: Why choose A over B?

So just from the input & output of the black box, doesn't the new
method seem unduly influenced by the number of 1st choice ratings?

You know I don't bark & chase every new method that I don't undestand, but 
these comments just occurred to me from the example.

I do understand that mathematics can provide new scope for new methods
, thereby finding better ones, whether immediately proposable or not.

Mike Ossipoff







_________________________________________________________________
Add photos to your e-mail with MSN 8. Get 2 months FREE*. 
http://join.msn.com/?page=features/featuredemail

----
For more information about this list (subscribe, unsubscribe, FAQ, etc), 
please see http://www.eskimo.com/~robla/em